Sepolia research identity

Agent 9481

The first EMR test agent: a public identity connected to reproducible task evidence, explicit authority limits, and transparent evaluation.

Network
Ethereum Sepolia
Standard
ERC-8004
Capability
BTC risk research
Authority
Simulation only

Experiment 001 · interactive prototype

Generate one execution receipt.

Choose a synthetic market snapshot. Agent 9481 applies a fixed, inspectable risk policy and produces a browser-generated evidence record. No live market data, funds, trades, or investment advice are involved.

01 Task input

Choose a frozen scenario

BTC spot
$100,000
30d realized vol.
38%
8h funding
0.005%
24h open interest
2.1%
25d put skew
1.2
BTC seven-day return
1.8%
02 Execution evidence
9481

No receipt yet

Run the fixed policy to generate a locally verifiable JSON record.

Ethereum activation path

Move from identity to evidence.

The identity is on-chain. The next steps connect it to public metadata, repeatable executions, independent evaluation, and eventually grounded ERC-8004 reputation.

  1. 01
    Register identity

    Agent 9481 was minted on the Sepolia ERC-8004 registry.

    Complete
  2. 02
    Publish metadata

    The registration JSON is publicly hosted and ready to attach.

    Ready
  3. 03
    Set the on-chain agent URI

    Call setAgentURI(9481, metadataUrl) from the owner wallet.

    Next
  4. 04
    Commit an execution record

    Download a receipt from this lab, review it, and add it to the research repository.

    Pending
  5. 05
    Add independent evaluation

    A second wallet or evaluator verifies the task evidence and declared method.

    Pending
  6. 06
    Submit grounded reputation

    Reference the evaluated receipt in an ERC-8004 reputation transaction.

    Pending

What this proves

Identity tells us who the agent is. An execution receipt begins to show what it did, under which limits, and how another evaluator can verify it.

This prototype validates the evidence pipeline—not Agent 9481's financial expertise. Expertise requires repeated tasks, delayed outcomes, counterfactual baselines, and independent evaluation.